bioRxiv preprint doi: https://doi.org/10.1101/2020.08.07.242388; this version posted August 10, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. The windblown: possible explanations for dinophyte DNA in forest soils Marc Gottschlinga, Lucas Czechb,c, Frédéric Mahéd,e, Sina Adlf, Micah Dunthorng,h,* a Department Biologie, Systematische Botanik und Mykologie, GeoBio-Center, Ludwig- Maximilians-Universität München, D-80638 Munich, Germany b Computational Molecular Evolution Group, Heidelberg Institute for Theoretical Studies, D- 69118 Heidelberg, Germany c Department of Plant Biology, Carnegie Institution for Science, Stanford, CA 94305, USA d CIRAD, UMR BGPI, F-34398, Montpellier, France e BGPI, Université de Montpellier, CIRAD, IRD, Montpellier SupAgro, Montpellier, France f Department of Soil Sciences, College of Agriculture and Bioresources, University of Saskatchewan, Saskatoon, S7N 5A8, SK, Canada g Eukaryotic Microbiology, Faculty of Biology, Universität Duisburg-Essen, D-45141 Essen, Germany h Centre for Water and Environmental Research (ZWU), Universität Duisburg-Essen, D- 45141 Essen, Germany Running title: Dinophytes in soils Correspondence M. Dunthorn, Eukaryotic Microbiology, Faculty of Biology, Universität Duisburg-Essen, Universitätsstrasse 5, D-45141 Essen, Germany Telephone number: +49-(0)-201-183-2453; email: [email protected] bioRxiv preprint doi: https://doi.org/10.1101/2020.08.07.242388; this version posted August 10, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. ABSTRACT Dinophytes are widely distributed in marine- and fresh-waters, but have yet to be conclusively documented in terrestrial environments. Here we evaluated the presence of these protists from an environmental DNA metabarcoding dataset of Neotropical rainforest soils. Using a phylogenetic placement approach with a reference alignment and tree, we showed that the numerous sequencing reads that were assigned to the dinophytes did not associate with taxonomy, environmental preference, nutritional mode, or dormancy. All the dinophytes in the soils are most likely windblown dispersal units of aquatic species, and are not biologically active residents of terrestrial environments. Keywords dinoflagellates; distribution; phylogenetic placement; operational taxonomic unit; rRNA. bioRxiv preprint doi: https://doi.org/10.1101/2020.08.07.242388; this version posted August 10, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Environmental high-throughput sequencing (HTS) studies of protists have now been performed for over a decade (Santoferrara et al. 2020). During that time, a large diversity of dinophyte DNA sequences has also been uncovered. Dinophytes are an ecologically and economically important group of protists that exhibit many types of life styles and nutritional modes, including phototrophic, mixotrophic and heterotrophic forms as well as some being parasitic (Saldarriaga and Taylor 2017). All known dinophytes are from marine or freshwater environments (Adl et al. 2019). As they constitute a considerable fraction of the plankton and play an important role in the global aquatic ecosystem, HTS studies have detected dinophytes from waters sampled from the polar regions through to the tropics (de Vargas et al. 2015; Le Bescot et al. 2016; Elferink et al. 2017; Decelle et al. 2018; Lentendu et al. 2018; Annenkova et al. 2020; Giner et al. 2020; Gottschling et al. 2020). HTS studies have also detected DNA of dinophytes in terrestrial environments (Bates et al. 2013; Geisen et al. 2015; Mahé et al. 2017; Venter et al. 2017; Voss et al. 2019), although they are not expected to be there. Aquatic protists can sometimes be detected in terrestrial environments, notably riparian soil, such as foraminifera (Meisterfeld et al. 2001; Lejzerowicz et al. 2010) and possibly haptophytes (Mahé et al. 2017). However, that does not mean the normally aquatic protists are biologically active in soils or other drier environments (Geisen et al. 2018). In the absence of observing putative soil dinophytes using direct microscopic observations, here we used Mahé et al.’s (2017) metabarcoding data from three lowland Neotropical rainforest soils to ask if the presence of dinophytes in those soils associate with taxonomy, environmental preference, nutritional mode, or dormancy. bioRxiv preprint doi: https://doi.org/10.1101/2020.08.07.242388; this version posted August 10, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. MATERIALS AND METHODS Environmental sampling and data Sampling and sequencing of tropical soils originally took place in lowland rainforest in Costa Rica, Panama, and Ecuador (Mahé et al. 2017). The extracted soils DNAs were amplified for the hyper-variable V4 region of the SSU-rRNA locus using general eukaryotic primers (Stoeck et al. 2010); this short region has relatively strong phylogenetic signal, although it is not as strong as the full-length SSU-rRNA (Dunthorn et al. 2014; Gottschling et al. 2020). Illumina sequencing reads were clustered into OTUs using Swarm v2 (Mahé et al. 2015) and taxonomically assigned to the Protist Ribosomal Reference database (Guillou et al. 2013) using VSEARCH (Rognes et al. 2016). The 269 OTUs that were assigned to the dinophytes by Mahé et al. (2017), were extracted and used here for phylogenetic placements (File S1). Reference tree From GenBank, 228 ingroup dinophytes, plus 10 outgroups, were downloaded, then aligned with MAFFT v6.624b (Katoh and Standley 2013) using the –auto option. Based on previous analyses (Gottschling et al. 2012, 2020; Žerdoner Čalasan et al. 2019), the full sequences of each species were used without excluding ambiguously aligned positions sites. Phylogenetic inferences of the reference alignment were carried out by using Maximum Likelihood (ML) as described in detail by Gottschling et al. (2012, 2020), using RAxML v8.2.10 (Stamatakis 2014) with the GTR+G substitution model. To determine the best fitted ML tree, we executed 10-tree searches from distinct random stepwise addition sequence Maximum Parsimony starting trees and performed 1,000 non-parametric bootstrap replicates. Reference alignment and tree available upon request. Phylogenetic placement of environmental OTUs The OTU representative sequences obtained from Swarm were aligned against the reference alignment using PaPaRa v2.0 (Berger and Stamatakis 2011), and phylogenetically placed onto the ML reference tree using the Evolutionary Placement Algorithm (EPA) of RAxML. Next, all OTUs that were placed with at least 95% probability (combined likelihood weight ratios) in the dinophyte clade were extracted and visualized, using Gappa (Czech et al. 2020). For details on the extraction, see Czech et al. (2018); details of the workflow are published in a GitHub code repository (https://github.com/lczech/dinoflagellate-paper). bioRxiv preprint doi: https://doi.org/10.1101/2020.08.07.242388; this version posted August 10, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. RESULTS AND DISCUSSION For the dinophyte reference alignment and tree, we included a broad representative taxon sample covering the known DNA sequence diversity with comprehensive sequence information. The alignment was 7,270 bp long and had 3,753 parsimony informative sites (52%, 15.7 per terminal taxon). The ML tree had many bipartitions that had high if not maximal bootstrap values. The Dinophyceae was inferred to be monophyletic, and it contained well-known subclades: Dinophysales, Gonyaulacales, Gymnodiniales, Peridiniales, Prorocentrales, †Suessiales as well as Amphidomataceae, Brachydiniaceae, and Tovelliaceae. Only 207 Neotropical soil OTUs from from the Mahé et al.’s (2017) that were assigned to the dinophytes, phylogenetically placed across the reference tree with high likelihood weight scores (Fig. 1, File S2). There were no exclusive associations with taxonomy. Some of the dinophyte OTUs formed distinct clades of sequences that were unknown until the present study. However, the amount of such undescribed diversity is low compared to other microbial lineages such as the Fungi (Jones et al. 2011; Rosling et al. 2011). They placed onto early branches, which comprise heterotrophic, mainly parasitic species (Saldarriaga et al. 2003; Gómez et al. 2009; Bachvaroff et al. 2012; Gu et al. 2013), but also within the Peridiniales. Most of the OTUs, though, placed within already known lineages of the Gymnodiniaceae, Peridiniales and †Suessiales, a truly heterogeneous set of dinophytes including unarmored and thecate algae as well. There is no morphological trait that the OTUs would therefore